This study has several strengths. It included a relatively large number of patients, reducing the possibility of bias by “masking” which treatment patients received, classifying participants according to the severity of their tinnitus and using highly standardised interventions. Also, the researchers used established scales to measure the severity of tinnitus and its impact on quality of life.
It is very well accepted that tinnitus often is "centralized" -- while it is usually initiated with an inner ear event, persistent tinnitus is associated with changes in central auditory processing (Adjamian et al, 2009). Sometimes this idea is used to put forth a "therapeutic nihilism" -- suggesting that fixing the "cause" -- i.e. inner ear disorder -- will not make the tinnitus go away.   This to us seems overly simplistic -- while it is clear that the central nervous system participates in perception of sounds, and thus must be a participant in the "tinnitus" process, we think that it is implausible that in most cases that there is not an underlying "driver" for persistent tinnitus.
A common cause of tinnitus is inner ear hair cell damage. Tiny, delicate hairs in your inner ear move in relation to the pressure of sound waves. This triggers cells to release an electrical signal through a nerve from your ear (auditory nerve) to your brain. Your brain interprets these signals as sound. If the hairs inside your inner ear are bent or broken, they can "leak" random electrical impulses to your brain, causing tinnitus.